Summary
Shobhita Sundaram is a PhD student and graduate research assistant at MIT CSAIL specializing in deep learning and large model research, with eight years of industry and research experience spanning Meta, Google Research, DeepMind, Apple, Two Sigma, and D. E. Shaw. Her work focuses on representation learning, data selection for pretraining, and self-improvement mechanisms for LLMs, blending rigorous academic inquiry with production-minded experimentation. She has contributed to vision and multimodal research (VisCam) and investigated long-range spatial relationships in neural networks during her undergraduate work at CBMM. Comfortable moving between research and engineering, she has built end-to-end systems from REST services to ML pipelines and applied reinforcement learning in quantitative settings. Based in Greater Boston, she maintains an accessible portfolio of her publications and code that showcases both theoretical depth and practical implementations.
8 years of coding experience
4 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at Massachusetts Institute of Technology